We are looking for a Product-First Senior AI Engineer who can bridge the gap between "Research" and "Reality." You won't just be importing libraries here; you will be architecting the intelligence engine that powers our Industrial AI platform. You will build high-performance Multimodal Systems (Vision + Language) that run efficiently on both Edge devices and the Cloud to solve complex logic problems for Fortune 500 supply chains.
Key Responsibilities
- Rapid Prototyping & Innovation: Don't just use open-source code—innovate on top of it. Experiment with the latest libraries to build novel solutions for complex logistics challenges.
- End-to-End Ownership: Translate abstract business requirements into logical, scalable AI architectures. You will own the solution from data analysis to model deployment.
- Performance Engineering: It’s not enough to be accurate; it must be fast. Diagnose, troubleshoot, and optimize inference pipelines for low latency and high throughput.
- Data-Centric AI: Drive the strategy for Data Analysis, Feature Engineering, and Augmentation. You will guide the annotation team and build pipelines to extract meaningful insights from massive Vision and Text datasets.
- Advanced Vision & NLP: Build robust solutions for Person/Scene understanding (Pose Estimation, Re-Identification) and integrate GenAI/LLM capabilities to add semantic understanding to visual data.
- Cross-Functional Collaboration: Work closely with the DevOps and Product teams to translate AI needs into effective, fault-tolerant technical solutions.
Skills & Requirements
- Production Python: Strong experience writing clean, modular, and fault-tolerant code. You understand that a model in a notebook is not a product.
- Deep Learning Stack: Proficiency in PyTorch is essential and experience with inference optimization tools like TensorRT is also required. Experience with practical edge deployement is a massive plus.
- Custom Model Training: Familiarity with training or fine-tuning custom AI models (Detectors, Classifiers) from scratch.
- Computer Vision Mastery: Deep understanding of Image Processing technologies (OpenCV, Dlib, NumPy) and modern architectures (YOLO, ResNet, etc.), OCRs and VLMs.
- NLP & GenAI: Hands-on experience with Hugging Face, LangChain, and NLP libraries (Spacy, NLTK). Ability to implement RAG pipelines or Agentic workflows.
- Complex Vision Tasks: Experience with advanced problems like Person Re-Identification, Pose Estimation, and Tracking.
- Applied AI: A proven track record of successfully applying machine learning to solve real-world problems (not just Kaggle competitions).
Brownie Points
- Cutting-Edge Tech: Knowledge of the latest advancements in AI, especially Vision Transformers (ViTs), CLIP, and Multimodal LLMs.
- DevOps Awareness: Understanding of Docker and Git. You know how to containerize your application for deployment.
- Cloud/Edge: Experience deploying models on AWS or NVIDIA Jetson devices.
What We Offer
- Meritocracy: A candid startup culture where the best ideas win.
- The Playground: Access to the latest NVIDIA Hardware and cutting-edge Generative AI tools.
- Ownership: Work with a performance-oriented team driven by autonomy and open to experiments.
- Impact: Design systems for high accuracy and scalability that physically move the global supply chain.
About KoiReader
KoiReader Technologies is a global MNC defining the future of Industrial AI. Founded by industry experts and powered by a strategic partnership with NVIDIA, we turn complex operational challenges into intelligent competitive advantages for our Fortune 500 clients.
Through KoiVision®—our patented Digital Operations Platform and AI Operating Model—we are setting the global standard for how Generative AI and Computer Vision automate supply chain and logistics. Craft your own destiny. Join the leaders of the AI Revolution.